Run on CPU?

[DecentMakeover]

Hi

Thanks for sharing Your work, I was wondering if i could run this only on the CPU?

[TreB1eN]

yes, you can. both CPU and gpu are supported

[JerryHanZD]

yes, you can. both CPU and gpu are supported

Hi, really thanks for your code! I'm working on a lab pc with no GPU, and trying to work on CPU. I met the problem "OSError: libcuda.so.1: cannot open shared object file: No such file or directory". Do I need to change the running environment? Or do you know how to solve this problem? Thx!!

[qingshan412]

Hi, really thanks for your code! I'm working on a lab pc with no GPU, and trying to work on CPU. I met the problem "OSError: libcuda.so.1: cannot open shared object file: No such file or directory". Do I need to change the running environment? Or do you know how to solve this problem? Thx!!

You can specify the map_location when restoring the existing model on different platforms, like self.model.load_state_dict(torch.load(save_path/'model_{}'.format(fixed_str), map_location=conf.device))

[ThanhPham1987]

Hi When i run model on CPU. I face issue: Illegal instruction (core dumped) can you give me some advice? Best regards, PeterPham

[AkaCoder404]

@ThanhPham1987 Hey I solve this issue by installing an earlier version of pip install mxnet==1.1.0 (when running python face_verify.py). Then installing the correct numpy. I'm not sure if this solve yours issue.